Fax: (011) 81 76-234-4252
Comparative analysis of epidermal growth factor receptor mutations and gene amplification as predictors of gefitinib efficacy in Japanese patients with nonsmall cell lung cancer
Version of Record online: 26 MAR 2007
Copyright © 2007 American Cancer Society
Volume 109, Issue 9, pages 1836–1844, 1 May 2007
How to Cite
Sone, T., Kasahara, K., Kimura, H., Nishio, K., Mizuguchi, M., Nakatsumi, Y., Shibata, K., Waseda, Y., Fujimura, M. and Nakao, S. (2007), Comparative analysis of epidermal growth factor receptor mutations and gene amplification as predictors of gefitinib efficacy in Japanese patients with nonsmall cell lung cancer. Cancer, 109: 1836–1844. doi: 10.1002/cncr.22593
- Issue online: 18 APR 2007
- Version of Record online: 26 MAR 2007
- Manuscript Accepted: 11 JAN 2007
- Manuscript Revised: 10 JAN 2007
- Manuscript Received: 30 OCT 2006
- epidermal growth factor receptor gene;
- gene amplification;
- fluorescence in situ hybridization;
Because the investigation of epidermal growth factor receptor gene (EGFR) status as a predictor of gefitinib efficacy in Japanese patients has shown promise, the authors evaluated EGFR mutations and gene amplification in biopsy specimens from Japanese patients with nonsmall cell lung cancer (NSCLC) who received treatment with gefitinib to analyze the correlation between EGFR gene status and clinical outcome.
Fifty-nine patients were enrolled in this study. EGFR gene amplification was evaluated by fluorescence in situ hybridization (FISH), and EGFR mutations in exons 18, 19, and 21 were analyzed by polymerase chain reaction and direct sequencing.
EGFR mutations were detected in 17 patients (28.8%). FISH-positive results were observed in 26 patients (48.1%). The response rate was significantly higher in the patients with EGFR mutations than in the patients without mutations (58.8% vs 14.3%; P = .0005). No significant difference in the response rate was observed between FISH-positive patients and FISH-negative patients (31.8% vs 21.4%; P = .4339). EGFR mutation was correlated with both a longer time to progression (TTP) (7.3 months vs 1.8 months; P = .0030) and longer overall survival (OS) (18.9 months vs 6.4 months; P = .0092). No significant differences in TTP or OS were observed between FISH-positive patients andFISH-negative patients. The results from a multivariate analysis indicated that EGFR mutations maintained a significant association with longer TTP and longer OS.
The results of this study suggested that EGFR mutations may serve as predictors of response and survival and that the role of EGFR gene amplification is not a predictor of gefitinib efficacy in Japanese patients with NSCLC. Cancer 2007. © 2007 American Cancer Society.
Epidermal growth factor receptor (EGFR) is a tyrosine kinase receptor that belongs to the ErbB family and has been implicated in cell proliferation and survival. The EGFR gene (EGFR) frequently is overexpressed in many solid tumors, including nonsmall cell lung cancer (NSCLC).1 Gefitinib (Iressa; Astra Zeneca) is a targeted agent that inhibits the tyrosine kinase activity of EGFR by competitively blocking the adenosine triphosphate binding site.2 Two Phase II trials (THE Iressa Dose Evaluation in Advanced Lung Cancer [IDEAL] 1 and IDEAL 2 studies) confirmed that gefitinib was active in from 10% to 20% of patients with NSCLC who had failed on standard therapy,3, 4 and both trials revealed significant variability in response to gefitinib. Objective responses to gefitinib frequently are observed in women, never-smokers, and patients with adenocarcinoma histology,3, 4 and molecular predictors of response to gefitinib have been investigated. Protein expression of EGFR and its downstream molecules have been studied widely immunohistochemically, but the results have not been consistent.5–8
In 2004, 3 groups reported that tumors with somatic mutations in the kinase domain of EGFR were associated with gefitinib sensitivity.9–11 Several retrospective studies have revealed that EGFR mutations are observed more frequently in women, never-smokers, patients with adenocarcinoma histology, and East-Asian patients12–17 and that mutations are associated with objective response to gefitinib and a prolonged time to progression (TTP). A survival benefit of gefitinib treatment has been observed in many East-Asian patients with EGFR mutations, but no such benefit has been reported in Caucasian patients with EGFR mutations.7, 12, 13, 15, 18, 19
Cappuzzo et al. investigated EGFR gene amplification by fluorescence in situ hybridization (FISH) in patients with NSCLC who received treatment with gefitinib, and their results indicated that EGFR gene amplification was associated with an objective response to gefitinib, a longer TTP, and a longer overall survival (OS).14, 18 Those authors also demonstrated that EGFR gene amplification assessed by FISH was a more reliable biomarker for predicting a survival benefit of gefitinib therapy in Caucasian patients with NSCLC than EGFR mutation status. However, it is unclear whether EGFR gene amplification assessed by FISH is an effective predictor of gefitinib efficacy in Japanese patients with NSCLC, who frequently have EGFR mutations and in whom the mutations seem to be reliable predictors of gefitinib efficacy. In the current study, we evaluated the EGFR mutation status and gene amplification in biopsy specimens from Japanese patients with advanced NSCLC to assess their predictive value in regard to the efficacy of gefitinib in this population.
MATERIALS AND METHODS
This was a retrospective study and was conducted at 5 institutions: Kanazawa University Hospital, Ishikawa Prefectural Central Hospital, Kouseiren Takaoka Hospital, Houju Kinen Hospital, and Kanazawa Municipal Hospital. Gefitinib therapy in the enrolled patients was started between July 2002 and October 2005. The patients were chosen for this study based on the following eligibility criteria: histologically confirmed NSCLC treated with gefitinib (250 mg per day), the ability to obtain tumor samples from each institution after receiving approval from the institution's review board and the patient's written informed consent, and the ability to obtain complete clinical information and physical examination records, including information on Eastern Cooperative Oncology Group performance status (PS).20
Patients who received treatment with gefitinib were evaluated for response every 4 weeks according to the Response Evaluation Criteria in Solid Tumors (RECIST) guidelines.21 Tumor response was assessed by computed tomography (CT), and a confirmatory evaluation was repeated after 4 weeks in patients who had a complete response (CR), a partial response (PR), or stable disease (SD). TTP was calculated from the date gefitinib therapy was started to the date of disease progression or the date of last contact. OS was calculated from the date gefitinib therapy was started to the date of death or the most recent date on which the patient was known to be alive. Some patients in this study had been enrolled in our previous study, which was a Phase II trial that evaluated the clinical benefit of gefitinib in chemotherapy-naive patients with NSCLC.22 The investigators who examined the biopsy specimens were blinded to the clinical outcome.
The tumor specimens were fixed in formalin and embedded in paraffin wax. Serial sections that contained representative malignant cells were deparaffinized in xylene washes and dehydrated in 100% ethanol. No specimens were microdissected.
DNA was extracted from 5 serial, 10-μm-thick sections by using the QIAamp DNA Mini kit (Qiagen, Tokyo, Japan) according to the protocol described in the manufacturer's instructions. The DNA obtained was eluted in 50 μL of buffer AE, and the concentration and purity of the extracted DNA were assessed by spectrophotometry. The extracted DNA was stored at −20°C until it was used. EGFR mutations were detected by polymerase chain reaction (PCR)-based direct sequencing of exons 18, 19, and 21. PCR amplification was performed in 100 ng of genomic DNA with the TaKaRa Ex Taq Hot Start Version kit (TaKaRa, Tokyo, Japan). The following primers were used: exon 18, 5′-CCTTGTCTCTGTGTTCTTGT-3′ (forward) and 5′-CTGCGGCCCAGCCCAGAGGC-3′ (reverse); exon 19, 5′-CATGTGGCACCATCTCACA-3′ (forward) and 5′-CCACACAGCAAAGCAGAAAC-3′ (reverse); and exon 21, 5′-CAGGGTCTTCTCTGTTTCAG-3′ (forward) and 5′-TAAAGCCACC TCCTTACTTT-3′ (reverse). DNA was amplified for 35 cycles at 95°C for 30 seconds, at 61°C for 30 seconds, and at 72°C for 60 seconds followed by 7 minutes of extension at 72°C. Sequencing was performed with a 3100 Genetic Analyzer (Applied Biosystems, Tokyo, Japan), and the results were analyzed by using Sequencer 3.11 software (Applied Biosystems, Foster City, Calif) to compare variations. The sequences were compared with the GenBank sequence for human EGFR (accession number AF288738).
Each slide was deparaffinized in xylene washes and dehydrated in 100% ethanol. After incubation in 0.2 N HCL at room temperature for 20 minutes and in a pretreatment reagent (NaSCN; Vysis, Tokyo, Japan) at 80°C for 30 minutes, the slides were digested with a proteinase reagent at 37°C for 10 to 60 minutes. Slides were refixed in 10% neutral buffered formalin at room temperature for 10 minutes and rinsed twice in 2 × standard saline citrate (SSC), pH 7.3, at room temperature for 5 minutes each. The slides were incubated in 70% formamide/2 × SSC, pH 7.0 to 8.0, at 72°C for 5 minutes to allow the chromosomes to denature and were then dehydrated in a series of increasing concentrations of ethanol solutions (70%, 85%, and 100%). The EGFR/chromosome 7 centromere (CEP7) probe set (Vysis) was applied to an area that was selected based on the presence of tumor foci on each slide, and the hybridization area was covered with a cover glass and sealed with rubber cement. The slides were incubated in a humidified chamber at 37°C for 20 to 24 hours to allow hybridization to occur. A posthybridization wash was performed in 2 × SSC/0.3% NP-40 at 73°C for 2 minutes and then in 2 × SSC at room temperature for 5 minutes. Finally, 4′,6-diamidino-2-phenylindole was applied to the target area, which then was covered with a coverslip.
The patients were classified into the following 6 categories according to EGFR gene copy numbers per cell and frequency of tumor cells with a specific copy number of the EGFR gene and CEP7, as described18 by Cappuzzo et al.: disomy (≤2 copies in >90% of cells), low trisomy (≤2 copies in ≥40% of cells, 3 copies in 10–40% of cells, ≥4 copies in <10% of cells), high trisomy (≤2 copies in ≥40% of cells, 3 copies in ≥40% of cells, ≥4 copies in <10% of cells), low polysomy (≥4 copies in 10–40% of cells), high polysomy (≥4 copies in ≥40% of cells), and high gene copy number (defined as the presence of tight EGFR gene clusters and a ratio of EGFR gene to chromosome of ≥2 copies or ≥15 copies of EGFR per cell in ≥10% of cells) (Fig. 1A-D). FISH analysis was performed independently by 2 investigators who were blinded to the patients' clinical characteristics and all other molecular variables. Patients with high polysomy or high gene copy numbers were classified as FISH-positive, and the remaining groups (with disomy, low trisomy, high trisomy, or low polysomy) were classified as FISH-negative.
Associations between response to gefitinib and other variables, including sex, histology, smoking history, EGFR mutations, and EGFR gene amplification, were analyzed for independence by using the chi-square test. A logistic regression model was used to identify which independent factors had a joint significant influence on the rate of objective response to gefitinib. TTP and OS were calculated by using the Kaplan-Meier method, and differences between patient groups according to EGFR mutations or FISH analysis were compared by using the log-rank test. A Cox proportional-hazards model was used for the multivariate analysis of survival. A P value of .05 was regarded as statistically significant, unless stated otherwise, and all comparisons were 2-sided. StatView 5.0 statistical software was used to perform all analyses.
The characteristics of the patients are shown in Table 1. Tumor samples were collected from 59 patients. We were able to obtain complete clinical information on 101 patients with NSCLC who received treatment with gefitinib at the 5 institutions. It was impossible to obtain tumor samples from 37 of 101 patients; because, in 22 patients, the diagnosis was made cytologically, and the 15 other patients already had been diagnosed at other hospitals before their admission to 1 of our institutions. Response to gefitinib was not evaluated in 5 patients, because they did not have lesions that could be evaluated.
|Variable||No. of patients||No. of responders||Response rate, %||P|
|EGFR gene amplification†|
In total, 59 tumor samples, 46 from primary tumors (29 obtained by transbronchial lung biopsy, 3 obtained by percutaneous lung biopsy, and 14 obtained at surgery) and 13 samples from lung tumor metastases to other sites (5 in lymph nodes, 5 in the brain, 1 in the small bowel, 1 in bone, and 1 in muscle) were obtained. All resections were performed for histologic diagnosis or palliative therapy. Twenty-four of 59 patients (40.7%) were women, and 21 patients (35.0%) were never-smokers. According to histologic type, there were 44 adenocarcinomas (73.3%; including 2 that contained bronchioloalveolar carcinoma components), 11 squamous cell carcinomas, 3 large cell carcinomas, and 1 adenosquamous cell carcinoma. Thirty-six patients (61.0%) had a good PS (0 or 1). Fifty-four patients (91.5%) had stage IV disease at the start of gefitinib therapy.
Twenty-five patients had been treated with chemotherapy, and 34 patients were chemotherapy-naive. A PR was achieved in 16 patients (27.1%), 20 patients had SD (33.9%), and 23 patients had progressive disease (PD) (39.0%). The overall response rate was 27.1% (95% confidence interval, 15.8–38.5%), and the response rate was significantly higher among women, patients with adenocarcinoma, and never-smokers (Table 1).
Fifteen patients were alive at the time of the analysis, and 5 of those patients still were receiving gefitinib therapy without tumor progression. The median follow-up was 9.9 months, the median TTP was 3.1 months (range, 0.2–34.8 months), and the median OS was 7.6 months (range, 0.7–38.8 months). The median duration of gefitinib therapy was 3.5 months. The TTP and OS were significantly longer among women (TTP: 4.3 months vs 1.1 months; P = .0084; OS: 18.2 months vs 6.4 months; P = .0022), among patients with adenocarcinoma (TTP: 3.7 months vs 1.1 months; P = .0024; OS: 10.2 months vs 6.1 months; P = .0025), and among never-smokers (TTP: 6.5 months vs 3.3 months; P = .0013; OS: 18.2 months vs 6.1 months; P = .0003). There were no differences in TTP or OS between patients who received gefitinib as first-line therapy and patients who received gefitinib as second-line therapy (TTP, 3.3 months vs 2.6 months, respectively; P = .3287; OS: 8.2 months vs 6.2 months, respectively; P = .2689).
EGFR mutations of exons 18, 19, and 21 were analyzed in every patient were detected in 17 patients (28.8%). The mutations consisted of in-frame deletions in exon 19 (n = 11 patients), an in-frame deletion with point mutation in exon 19 (n = 1 patient), and point mutations in exon 21 (n = 5 patients). EGFR mutations were significantly more frequent among women (12 of 24 women; 50.0%; 5 of 35 men, 14.3%; P = .0028) and never-smokers (10 of 21 never-smokers; 47.6%; 7 of 38 smokers; 18.4%; P = .0177). Mutations frequently were detected in adenocarcinomas (14 of 44 tumors; 31.8%); they were detected in 2 squamous cell carcinomas and in 1 large cell carcinoma, but none of the differences according to histologic subtype were statistically significant (Table 2).
|Variable||No. of patients with EGFR mutations||%||P||No. of FISH-positive patients||%||P|
EGFR gene amplification was assessed by FISH in 54 patients. The tissue specimens in the other 5 patients were small and inadequate for FISH analysis. Disomy for the EGFR gene was present in 5 patients (9.3%), low trisomy was present in 5 patients (9.3%), high trisomy was present in 0 patients (0%), low polysomy was present in 18 patients (33.3%), high polysomy was present in 5 patients (9.3%), and high gene copy numbers were present in 21 patients (38.9%). Twenty-six patients (48.1%) were FISH-positive, and 28 patients (51.9%) were FISH-negative. FISH-positive results were observed frequently among patients with adenocarcinomas and among never-smokers, but the differences were not statistically significant. FISH status was not associated with sex (Table 2). FISH-positive results were observed in 62.5% of patients with EGFR mutations (10 of 16 patients) and in 42.1% of patients without EGFR mutations (16 of 38 patients). The rate of FISH-positive results was not correlated with the presence of EGFR mutations (P = .1708). The clinical characteristics of the patients with EGFR mutations and their classification according to EGFR gene amplification in are summarized in Table 3.
|Patient||Sex||Histology||Smoking status||Response to gefitinib||Median OS, mo||Median TTP, mo||EGFR gene amplification||EGFR mutation|
|28||W||ADC||Smoker||SD||14.4||14.4||LP||del L747-S752, E746I|
EGFR Mutations and Gene Amplification As Predictors of Gefitinib Efficacy
The response rate among patients with EGFR mutations was significantly higher than the response rate among patients without mutations (58.8% vs 14.3%; P = .0005) (Table 1), but there was little difference in the response rate between FISH-positive patients and FISH-negative patients (30.8% vs 21.4%, respectively; P = .4339) (Table 1). A multivariate analysis of response to gefitinib that included the variables sex, histology, smoking status, and EGFR gene status revealed that EGFR mutation was the only factor that contributed significantly toward a response to gefitinib (P = .0196) (Table 4).
|Variable||OR for response||P||HR for TTP||P||HR for OS||P|
|EGFR gene amplification|
TTP and OS were significantly longer in patients with EGFR mutations than in patients without EGFR mutations (TTP: 7.3 months vs 1.8 months; P = .0030) (Fig. 2A) (OS: 18.9 months vs 6.4 months; P = .0092) (Fig. 2B), but no significant difference in the median TTP or OS was observed between FISH-positive patients and FISH-negative patients (TTP: 1.8 months vs 2.6 months, respectively; P = .8535) (Fig. 2C) (OS: 6.4 months vs 8.2 months, respectively; P = .9121) (Fig. 2D). Five variables (sex, histology, smoking status, EGFR mutation, and EGFR FISH-positive results) were included in a Cox hazards model to define which variables were predictive of longer survival. The results showed that EGFR mutation was the only factor associated with both longer TTP and longer OS (Table 4). The median OS of patients with mutations in exon 21 was longer than of patients with exon 19 deletion mutations (23.0 months vs 18.2 months; P = .0271), but there was no difference in response rate or TTP between patients with mutations in exon 19 and patients with mutations in exon 21 (response rate: 60.0% vs 58.3%, respectively; P = .9493; TTP: 8.3 months vs 6.4 months, respectively; P = .5158). TTP and OS were significantly longer in the 10 responders with mutations than in the 6 responders without mutations (TTP: 15.3 months vs 3.3 months; P = .0022; OS: 20.0 months vs 7.7 months; P = .0007).
The results of this study showed that EGFR mutation was the only factor significantly associated with a better response to gefitinib, longer TTP, and longer OS in Japanese patients with NSCLC. A multivariate analysis revealed that EGFR mutation was a more reliable predictor of gefitinib efficacy than sex, histology, smoking status, or EGFR gene amplification. All biopsy specimens were examined successfully for EGFR mutations, and the results were comparable with those from previous studies in which surgical specimens were used,12, 17, 19 even though the amount of tumor tissue available in the biopsy specimens was limited. The results suggested that gefitinib efficacy in Japanese patients with NSCLC can be predicted by detecting EGFR mutations in biopsy specimens.
Several recent studies produced higher response rates in patients with mutations in exon 19 of EGFR than in patients with L858R in exon 21 and demonstrated that TTP and OS were longer in patients with the deletion than in the patients with L858R.23, 24 However, our current results showed no difference in response or TTP between patients with exon 19 mutations and patients with exon 21 mutations, and patients with the L858R point mutation in exon 21 had a longer median OS than patients with a point mutation in exon 19. We cannot explain why our results were not in line with the those obtained in previous studies.23, 24 Our study was not designed investigate the differences in gefitinib efficacy between patients with mutations in exon 19 and patients with mutations in exon 21, and the sample size of the study was too small to detect a statistically significant difference.
The small number of patients is a clear limitation of this study. A large-scale, single-arm study of Western NSCLC reported that both EGFR mutations and gene amplification were associated with response rate to gefitinib, longer TTP, and longer OS, and that study also demonstrated a significant correlation between FISH-positive results and the presence of EGFR mutations in paired specimens.18 Results from samples in prospective, placebo-controlled, Phase III studies were published subsequently, and the usefulness of EGFR gene amplification as a predictor of gefitinib efficacy was validated in a Phase III, placebo-controlled study, the Iressa Survival Evaluation in Lung Cancer (ISEL) trial.25 The results of that prospective study indicated that high EGFR gene copy numbers were a predictor of a better response to gefitinib and a survival benefit.26 The results from the current study conflict with the results from the ISEL trial and do not demonstrate any association between gene amplification and the efficacy of gefitinib therapy.
One possible explanation for the discrepancies between findings from the studies described above and our own findings is the difference in EGFR gene status according to ethnicity. Han et al. investigated EGFR gene mutations, gene amplification, K-ras mutation, and Akt phosphorylation in tumor samples from East-Asian patients with NSCLC27 and demonstrated that EGFR mutation was an independent predictor of response and survival in a multivariate analysis that included EGFR gene amplification. Those authors also demonstrated that FISH-positive results were associated with a better response rate the same as EGFR mutation in the univariate analysis but was not associated with prolonged survival. Previous studies have shown an approximately 3 times higher incidence of EGFR mutations in East Asians than in Caucasians.7, 9–12, 16, 17, 19, 21, 28–30 FISH-positive results do not appear to contribute significantly to the response to gefitinib or to survival in populations with high percentages of EGFR mutations.
The incidence of L858R in our study seemed low compared with the incidence of exon 19 deletion. Previous studies have demonstrated that the incidence of deletion mutations in exon 19 is almost the same as the incidence of point mutations in exon 21.7, 9–12, 16, 17, 19, 21, 28–30 Because the direct sequencing method usually was used to detect EGFR mutations, it is unlikely that the low frequency of the L858R mutation was caused by assay-related, false-negative findings. Our results of the incidence of L858R mutation and exon 19 deletion mutations may also produce some distortion with regard to the analysis of gene copy numbers because the number of patients with high gene copy numbers has been observed to be higher in those with deletion mutations in exon 19 than with point mutations in exon 21. Further analyses in much larger groups of patients will be necessary to clarify the frequency of the 2 most common mutations.
Takano et al. demonstrated an association between increased EGFR copy numbers measured by quantitative PCR (qPCR) and both higher a response rate and longer TTP.15 Dziadziuszko et al. reported that EGFR messenger RNA (mRNA) expression in tumor samples measured by qPCR was a predictive biomarker for response to gefitinib and longer progression-free survival. Those investigators also demonstrated that EGFR mRNA expression measured by qPCR was correlated significantly with FISH-positive results.31 It is possible that qPCR may enable us to make a more reliable distinction between specific and nonspecific amplification of the EGFR gene.13 We plan to compare EGFR gene copy numbers in corresponding samples measured with qPCR to confirm our results. We classified all patients into a FISH-positive group and a FISH-negative group according to the scoring system published by Cappuzzo et al. EGFR gene copy numbers also may vary according to ethnicity, similar to the differences in EGFR mutation frequency. The definition of FISH-positive results may need to be modified to use it as a predictor of gefitinib efficacy in Japanese patients with NSCLC.
In conclusion, the results of the current study suggest that the presence of EGFR mutations detected in biopsy specimens is an independent and significant predictor of response to gefitinib and survival in Japanese patients with advanced NSCLC. However, the role of EGFR gene amplification was not identified as a predictor of gefitinib efficacy in Japanese patients. Precise measurements are needed, and the validity of the classification must be confirmed in a prospective study.
We thank TSL (Tokyo, Japan) for suggesting that they use fluorescence in situ hybridization to test for epidermal growth factor receptor gene amplification.
- 7Epidermal growth factor receptor (EGFR) downstream molecules as response predictive markers for gefitinib (Iressa, ZD1839) in chemotherapy-resistant non-small cell lung cancer. Int J Cancer. 2005; 113: 109–115., , , et al.
- 8Sensitivity to gefitinib (Iressa, ZD1839) in non-small cell lung cancer cell lines correlates with dependence on the epidermal growth factor (EGF) receptor/extracellular signal-regulated kinase 1/2 and EGF receptor/Akt pathway for proliferation. Mol Cancer Ther. 2004; 3: 465–472., , , et al.
- 25Gefitinib plus best supportive care in previously treated patients with refractory advanced non-small-cell lung cancer: results from a randomised, placebo-controlled, multicentre study (Iressa Survival Evaluation in Lung Cancer). Lancet. 2005; 366: 1527–1537., , , et al.